Two different things wearing the same green label
A lot of regenerative agriculture marketing collapses two distinct concepts into one claim: "this practice reduces emissions." In accounting terms, reducing emissions and sequestering carbon are not the same operation, and a Scope 3 inventory has to treat them differently.
An emissions reduction means less greenhouse gas enters the atmosphere in the first place — less synthetic fertilizer applied, less diesel burned for tillage passes, less enteric methane per unit of output. This is a flow you can subtract from a baseline within the same accounting category.
Carbon sequestration is different. It means carbon that would otherwise be in the atmosphere gets pulled out and stored in soil organic matter or plant biomass. That's a removal, not a reduction, and it comes with a problem reductions don't have: reversibility. Fertilizer you didn't apply stays unapplied. Carbon stored in soil can be released again by a single tillage event, a drought, a land-use change, or simply time, if the practice that stored it stops.
What the GHG Protocol actually asks for
The GHG Protocol's separate guidance on land sector and removals accounting exists specifically because the Corporate Value Chain (Scope 3) Standard was not built with soil carbon dynamics in mind. The core instruction that matters for practitioners: removals should be quantified and reported separately from gross emissions reductions, not netted together into a single "reduction" number without clear labeling of what's actually happening.
For a company buying agricultural commodities, this shows up almost entirely in Category 1, Purchased Goods & Services — the supplier-facing category where farm-level practices enter the inventory. A supplier switching to reduced tillage or cover cropping may lower the emissions intensity of the crop (a reduction) while also building soil carbon stock over several seasons (a removal). Both can be real. Reporting them as one undifferentiated number is not.
Why soil carbon is not like a fuel switch
When a supplier switches from diesel to electric equipment, the accounting is comparatively simple: an activity that used to emit at rate X now emits at rate Y, and the difference is attributable to a specific, ongoing operational change. Soil carbon claims fail that simplicity on three counts:
- Baselines are contested. Soil organic carbon varies naturally by soil type, prior land use, and weather, so establishing what would have happened without the practice change requires either long time series or comparable control plots — something most supply chain programs don't have.
- Additionality is hard to demonstrate. If a farm was already building soil carbon through practices adopted for yield or cost reasons unrelated to a buyer's program, crediting that carbon to the buyer's Scope 3 inventory overstates the buyer's impact.
- Permanence has no fixed answer. Stored soil carbon needs monitoring across years, not a single measurement at program launch, because storage can be undone by a management change well after the reporting period closes.
None of this means soil carbon claims are illegitimate. It means they carry accounting obligations — monitoring commitments, uncertainty disclosure, reversal risk — that a straightforward emissions-reduction claim does not.
The measurement problem in practice
This is where the Scope 3 data quality hierarchy becomes directly relevant. Most regenerative agriculture claims currently entering corporate inventories rely on modeled or average data: a practice is adopted, and an emissions factor or sequestration coefficient from a model or published study is applied to the acreage involved. That's a legitimate starting point, but it sits well below supplier-specific primary data on the quality hierarchy, and it is a poor basis for a public removal claim specifically, because models are calibrated on average conditions that may not reflect the field in question.
Moving up the hierarchy for soil carbon means direct soil sampling at set depths, repeated over multiple years, ideally with paired control sites. That's expensive and slow relative to the pace at which sustainability claims get published. The honest position for a company reporting on supplier regenerative programs is to say clearly which figures are modeled estimates and which are measured, and to avoid presenting either as a settled, permanent reduction to gross inventory totals.
Where this discipline gets tested
As disclosure regimes mature, the gap between marketing claims and defensible accounting narrows fast. CSRD's ESRS E1 requires disclosure of removals and carbon storage separately from gross emissions figures, under a framework of double materiality that expects methodology and uncertainty to be stated, not just headline numbers. California's SB 253 brings phased third-party assurance to Scope 3 disclosures for large companies doing business in the state, which means removal and reduction claims embedded in a Scope 3 inventory will eventually face the same scrutiny as fuel and energy data. A soil carbon claim that can't show its baseline, its measurement method, and its monitoring plan will not hold up under assurance, even if the underlying agronomy is sound.
The practical takeaway
Before a regenerative agriculture claim goes into a Scope 3 inventory or a public report, it should be possible to answer three questions: is this a reduction or a removal, what data quality tier supports the number, and what happens to the claim if the practice lapses next season. Companies that can answer all three are in a defensible position. Companies that can't are making a marketing statement, not an accounting one.